ledc-dino
Library for discovery of inconsistencies and outliers
Description:
A concise repository that bundles algorithms for data-driven discovery of errors in a dataset. The main formalism is that of tuple-level constraints expressed as relational algebraic selection criteria.
More Information:
Originates from the AI-Driven Data Science research challenge (FAIR 2019–2023).
Responsible Research Lead: Guy De Tré
Authors: Antoon Bronselaer, Toon Boeckling, Milan Peelman
Programming Language: Java
Categories: Relational databases, anomaly detection
Publications:
A. Bronselaer, T. Boeckling, F. Pattyn, "Dynamic repair of categorical data with edit rules", Expert Systems with Applications, vol. 201, pp. 117-132, 2022.
M. Peelman, A. Bronselaer, G. De Tré, "Discovery of pairwise ordinal edit rules", In Joint Proceedings of the 19th World Congress of the International Fuzzy Systems Association (IFSA), the 12th Conference of the European Society for Fuzzy Logic and Technology (EUSFLAT), and the 11th International Summer School on Aggregation Operators (AGOP), pp. 109-116, 2021.
T. Boeckling, A. Bronselaer and G. De Tré, "Mining data quality rules based on T-dependence", In Proceedings of the 11th Conference of the European Society for Fuzzy Logic and Technology (EUSFLAT 2019), pp.184-191, 2019.